Post-stack fracture prediction technique based on improved phase decomposition

Abstract Amplitude variation with azimuth methods have a huge advantage in predicting small-scale fractures. However, the technique of using post-stack seismic data to predict fractures has rarely been investigated, with the main limiting factor being the degradation of resolution due to the stacking process, which makes it difficult to effectively identify small fractures. With the exploration and development of oil and gas fields entering into deeper levels, the traditional post-stack fracture prediction techniques often fail to meet the requirements of fine development. In this study, we propose a post-stack fracture enhancement technique based on phase decomposition. The technique advances parity phase decomposition based on the matched pursuit algorithm to extract the dominant phases in the seismic record, and the filtering process is carried out by the bidimensional empirical mode decomposition, which in turn computes the improved post-stack coherence and similarity attributes. Applying this method to 3D seismic datasets in the Sichuan region, the results indicate that the approach improves the seismic data quality and enhances the seismic attribute identification of fractures. The phase decomposition-based post-stack fracture enhancement technique shows significant potential in improving the signal quality of multidimensional seismic features such as texture, dip, and likelihood compared to traditional methods. This new technique provides a new way for fracture detection and characterization using post-stack seismic data, which can help to meet the needs of fine oilfield exploration and development.